Organizing knowledge the Chinese way



This paper examines an approach to knowledge organization that is fundamentally different from the analytic model evolved from ancient Greek philosophy. The classification being studied is the scheme used in a catalog, entitled the Seven Epitomes (Qilue), for organizing the Chinese imperial library collection in the Former Han dynasty over 2,000 years ago. Its knowledge organization model influenced almost all bibliographic classifications throughout imperial China. By applying a multidimensional framework to place the target scheme in its own cultural and historical context, the study identifies and discusses three key aspects of the Chinese classificatory approach. First, the scheme purposefully embraces a ru classicist ideology as its guiding principle. Second, it presents the knowledge universe as an integrated whole. Third, it applies correlative, rather than analytic, thinking to organizing textual categories, with the so-called ru Classics in the center of all major relationships. In conclusion, suggestions are made to identify additional aspects of the Chinese tradition in knowledge organization, to incorporate this way of thinking and test how it can work in constantly evolving information environments, and to understand categories and relationships in cultural contexts.


Mainstream, if not all, bibliographic classifications of today invariably take a fundamentally analytic approach that has evolved from ancient Greek philosophy. At its root, this analytic approach requires identification of concepts and their characteristics said to emerge from objective examination and or observation. Concepts that share a common set of characteristics are considered to belong to the same category; categories that share a common set of characteristics belong to the same broader category, and so forth. The focus of the approach is individual concepts, their characteristics, and how concepts relate to or differ from one another through their similar and different characteristics. Having had exclusive experiences with classifications of the type, the majority of us seem to be under the presumption that this is the only approach to knowledge organization, at least the only one that makes sense and works. Is that so?

The simplest answer to that question is an affirmative no. In this paper, I will describe part of the findings regarding a remarkably different approach to knowledge organization taken by a Chinese classificationist, Liu Xin, in his library catalog, the Seven Epitomes (Qilue), two-thousand years ago. My intention is not to suggest that a scheme created in a distant time and culture could be adopted for use in modern society. Rather, it is the unique epistemological approach to structuring knowledge in the target scheme to which I am paying attention. That approach influenced all other traditional Chinese bibliographic classifications for two millennia. For the most part, the model is said to have functioned well and achieved the intended purposes in a culture that highly prided itself on its writing and intellectual heritage (Fu, 2007). Such an effective approach to organizing knowledge is no doubt worth exploring, for the findings from this exploration will widen the horizon in classification research by both showing an alternative and illuminating the fallacy of universality.

The following includes first a brief review of the literature on the classification applied in the Seven Epitomes. Next is a description of the study's research methods as well as some challenges in the project. In the findings section, I will report on several aspects of the classification emerging as part of the investigation. The discussion centers around three themes of Chinese thinking manifested in the catalog's main structure. As a report on early findings, this paper concludes with a number of suggestions for future research.


The Seven Epitomes and Its Classification

As both the Seven Epitomes and its classification are surely unfamiliar to the ASIST participants, it seems necessary to introduce them before discussing the literature. What is described here is much abbreviated; more information about the Seven Epitomes is available in Lee & Lan (2009). The Seven Epitomes was the first documented library catalog in China's history, compiled right before the beginning of the Common Era. In 26 BCE, a distinguished scholar and government official, Liu Xiang (77–6 BCE),11 received an order from the emperor to head a collation project that would last for over 20 years and eventually create an organized imperial library. His son, Liu Xin (53 BCE-23 CE), was the one who completed the project and the classified catalog. Although the catalog is no longer extant in its original form, important historical writings of the time and a little later give several similar accounts of the project and provide pivotal details about the catalog. A commonly accepted belief in Chinese bibliography is that the History of the Former Han Dynasty (Han shu), a currently extant and important historical document composed at the end of the first century, in fact includes an abbreviated and slightly modified version of the catalog as its 30th chapter, the Bibliographic Treatise (yiwen zhi; also known as the Han Treatise).22 We are thus able to see the 6 main classes (i.e., 6 epitomes, with an extra epitome titled the “Collective Epitome” that is not a class),33 38 subclasses, and almost all entries (significantly shortened) in the catalog originally. Table 1 displays the basic structure of its classification including the main classes and subclasses. A more detailed explanation of this structure is given in the next section titled “Approach to Knowledge Organization.”

Related Writings in English

Traditional Chinese bibliographic classification has not attracted much notice in library and information science of the West with the exception of two historical overviews in English (Tsien, 1952; Jiang, 2007). However, the Seven Epitomes frequently appears in works on philosophy and intellectual history of early China. Most of those studies examine its treatment of the so-called Confucian Classics and categorization of the masters and propose new ways to align the thinkers and to revise intellectual history (e.g., Csikszentmihalyi & Nylan, 2003). Some others use its classification in the last two epitomes as a framework to assist with the understanding of recently excavated early texts in divination, natural history and medicine (e.g., Kalinowski, 2004). Without a doubt, classificatory approaches or methods are the least of their concern.

Table 1. The classification of the Seven Epitomes
Epitome (class)Subclass
Six ArtsChanges
 Spring and Autumn Annals
 Book of Filial Piety
MastersRu classicists
 Theorists of Yin-Yang
Lyrics and RhapsodiesRhapsodies 1
 Rhapsodies 2
 Rhapsodies 3
 Diverse rhapsodies
Military TextsMilitary Tactics
 Military Terrain
 Military Yin-Yang
 Military Skills
Divination and NumbersPatterns of Heaven
 Five Phases
 Milfoil and Turtle Shell
 Diverse Prognostications
 System of Forms
Formulae and TechniquesMedical Classics

Chinese Literature on Traditional Classification

Although it was deemed an indispensible guide to learning and scholarship, bibliographic classification itself was rarely a favorite for in-depth study throughout imperial China. Well-known traditional schemes seemed to all follow the model established by the Seven Epitomes with some adjustments in the main classes and subclasses (Fu, 2007). Generally speaking, no alternative approach to classification had been proposed. (It should be noted that this research is concerned with traditional Chinese classifications rather than the recent ones used in China and Taiwan that imitate modern American and European classifications.) A respected scholar in the 20th century goes so far as to assert that there was no bibliographic research despite of plentiful bibliographic tools created in the long Chinese imperial history; his comment is also applicable to the subfield of classification (Yu, 1963/2004). That means, systematic study of traditional classification appeared only in the recent 50 years or so, with the majority in the second half of that period. The following is a brief overview of this literature with special attention to the part relevant to the classification of the Seven Epitomes. Due to the limit of the paper, this review is unavoidably simplistic and sketchy. A more extensive and informative literature review can be found in Lee & Lan (in press).

In the Chinese literature on the target classification, a large portion, especially writings before the 1980's, is descriptive and or interpretive, embedded as a small segment within larger works on the Seven Epitomes in its entirety (e.g., Wang Y., 13th century/1983) or bibliography in general (e.g., Zhang X., 18th century/2002). This type of works helps modern-day researchers gain a basic understanding of the archaic scheme and its importance in bibliographic history. Another major strand of the discussion in the scholarly discourse is the catalog's ideological slant and its influence in and contribution to intellectual history, partly through its classification (Xu, 2009). On the other hand, a few authors criticize the catalog's ideology (Wang Z., 1984) and its “misplaced” preoccupation with intellectual heritages and developments (Wang G., 1991), pointing to its classification as a major problem. A small number of recent works have attempted to explain the origins of the main classes of the classification (e.g., Fu, 2003; Zhang S., 1994; Zou, 22008). Regrettably, these discussions rely solely on speculation and misinformation rather than rigorous research methods.

Overall, the classification in question, as well as the classificatory model it launched, remains an obscurity to date. Its theoretical approach and fundamental principles have eluded students of classification both in China and elsewhere.


Because the target classification is complex, we have planned a long-term research project to investigate its various facets that will require more than the space allotted here to detail. This paper narrowly reports on a preliminary examination of the general approach taken by the classification. But first, the following describes the research methods applied in the overall project.

Data in the project, all textual, comes from three major sources: (1) the Seven Epitomes (also referred to as the catalog in this paper), (2) early Chinese texts that are either recorded in the catalog or other texts similar to them, and (3) texts, mostly in Chinese with some in English, that provide the personal (as related to the classificationist), historical, social and technological context. The catalog itself serves as the primary source. As mentioned above, the catalog is no longer extant. At least 10 bibliographers in the Qing dynasty (1644–1912) individually attempted a reconstruction of it based on the Han Treatise and fragments quoted in many other sources. The version compiled by Yao Zhenzong (1843–1906) is considered to be the best, a newer edition of which is the basis for the current study (Liu X., 2008).

All data sources together form an analytical framework intended to address concerns and difficulties in all document studies as well as those unique in this particular case (Prior, 2003; Taylor & Bogdan, 1998). Such a multidimensional framework serves three pivotal purposes. First, a number of early texts can supplement authentic information related to the reconstructed catalog that consists of known gaps. Second, other texts will assist with contextualization of the catalog, thus helping to eliminate ahistorical and unwarranted interpretations of the catalog. Third, giving precedence to the catalog (i.e., it is considered more authoritative than others when there is contradiction) provides the analysis a genuine textual basis to avoid unsubstantiated speculation in the results.

Such a research framework arises from a firm belief that contexts are important for understanding knowledge organization systems. As Egan and Shera (1952) maintain, bibliography is part of social communication. Andersen and Skouvig (2006) echo this view by claiming that “knowledge organization is a social and political activity” (p. 316). To understand knowledge organization as such, it is thus methodologically critical to place the target classification in its own social, historical, political, and even technological context.


In this section, I will describe the target classification's structure and explain its six main classes and their relationships. This description then leads to a discussion of how the scheme manifests three general characteristics of Chinese thinking.

Classificatory Structure of the Seven Epitomes

The scheme applied for organizing entries in the Seven Epitomes is foreign to people trained in western ways of thinking. Even Chinese researchers in modern time often regard it as obsolete and irrelevant at best and unscientific and inferior at worst. Where in this classification is the analytic approach so ingrained in common schemes like Dewey Decimal Classification that we currently see in libraries around the world? Isn't it desirable to organize knowledge based on objective examination of the world rather than a subjective ideology? Why in the world did the classificationist, Liu Xin, choose the Classics, the writings by the masters, poetry, military texts, divination manuals, and medical texts to be the six main organizing themes? The translation of the six classes by Tsien (1952) is very revealing about the attempt by many to impose a western frame on something that is so unfamiliar: Classics, Philosophy, Poetry, Military Science, Science & Occultism, and Medicine. All of them match disciplines known today so naturally. But the match is reductionist and problematic, leading to criticisms that some categories seem to be exceptional or simply illogical.

Though there have been a number of interpretations of the catalog's classificatory structure in the literature, none seems credible or satisfactory. In a forthcoming article, our research team proposes a new interpretation that is based on a close and rigorous reading of the catalog itself within the framework described in the previous section on research methods (Lee & Lan, in press).

We suggest that the scheme in question comprises three ranked dichotomies, the impetus of which is ru classicism (commonly known as Confucianism, another reductionist translation, in the West). The first dichotomy is the one between the learning of dao (i.e., the Way [of living and thinking]) and the learning of qi (i.e., the vessel or practical skills). Texts in the “dao learning” camp are further dichotomized into the Classics (i.e., the six ancient texts held in the highest regards by classicists) and the non Classics. In the Classics category, interpretations of and commentaries on the Classics, two other important classicist texts and textbooks for foundational learning (e.g., texts to equip beginning students with necessary literary skills for tackling the Classics) also belong. The non Classics category then diverges into those that are expository and those not expository—thus the third dichotomy. Being less important, the “qi learning” camp includes technical texts connected to three types of government offices (those in charge of military, divination and healing) that became the three lesser categories. Each dichotomy resembles a binary opposition in which the two opposites are ranked. Classicism, the official ideology canonized by the Han throne since 136 BCE, establishes the preference between the opposites in ranking. Figure 1 illustrates the three sets of dichotomies, with the preferred category in each set on top, and how they form the six main classes in the scheme.

A Classification Driven by Ideology

There has always been a consensus on classicism being the guiding principle of the catalog's classification (e.g., Xu, 2009). Both Liu Xin and his father Xiang, the two leaders on the collation project that led to the creation of the catalog, were renowned classicists in their time when classicism became dominant in government and state-sponsored learning. It is not surprising that the catalog for the imperial library embraced the official canon.

Classicism is especially unequivocal in the catalog's classification. First of all, its foremost class is the Epitome of the Six Arts, referring to the six branches of learning centering on the Six Classics that form the textual basis of the official canon. Each of the first six subclasses in the Epitome represents one Classic (see Table 1). The seventh and eighth subclasses are devoted to two other texts, the Analects and the Book of Filial Piety, said to be teachings of Confucius who is the attributed founding figure in classicism (thus inspiring the convenient coinage of the term Confucianism). In the introductions to the other five epitomes and their various subdivisions, one or more quotations of the Classics or words of Confucius convey the classicist interpretations of or viewpoint toward other, non Classic text categories. For example, the introduction to the subclass “Diverse Prognostications” in the Epitome of Divination and Numbers cites the Book of Odes (the third Classic shown in Table 1) to attest to the importance of oneiromancy (i.e., divination by dreams) that precedes other types of divination in the subclass (Liu X., 2008, p. 105). Another evident sign of classicism's prominence in the classification is the placement of the category of “Classicists” as the lead among the 10 intellectual affiliations in the Epitome of the Masters.

Figure 2.

Main classes of the Seven Epitomes seen through three dichotomies.

This classicist overtone is no accident. After the short-lived first dynasty Qin (221–207 BCE), the Former Han (202 BCE-9 CE) was determined to make their empire last forever. Their empire-building efforts included intellectual control of which the institution of a literary canon was an indispensable part. The collation project that resulted in the creation of the imperial library and its catalog was intended to establish intellectual authority through an exertion of government control over writing and learning (Lewis, 1999). However, enshrining a particular group of texts or a particular group of thinkers radically departed from the censorship/book-burning policy of the Qin dynasty. The ideologically driven classification acted as the sole retrieval mechanism in the catalog and possibly in the imperial library, constraining the user to see the knowledge universe (the imperial library collection in this case) through a singular lens of classicism.

A Classification with a Holistic View

Sinologists often point to a holistic worldview (i.e., perceiving the world as an organic whole) in traditional Chinese philosophy (Liu, J., 2006). This cosmological stance was extended into all areas of thought in early China, including the view towards knowledge. Lewis (1999), for example, asserts that the Seven Epitomes “presumes unity of knowledge as the ideal …” (p. 326).

In its own text, the Seven Epitomes clearly imparts the idea that the Six Classics form a body of knowledge passed down from the sages that is the ultimate and supreme wisdom (Liu X., 2008). This knowledge must be pursued as a whole even though individual Classics embody various aspects of the whole. The catalog does not place an emphasis on elements/aspects of that supreme wisdom or on contrasts among the Classics. At the same time, Liu Xin explains that text contents in the other five categories are all fragmented, thus inferior to the Classics. Those other texts, nevertheless, are just fragments of the whole and can still return to the unity if a wise person can incorporate their merits and eliminate their weaknesses (Lewis, 1999). According to this point of view, the classification applied in the catalog is a unified scheme that has a focal class with five lesser, but unquestionably related classes.

It needs to be pointed out that Liu Xin never claimed his scheme to be exhaustive in its coverage. In effect, the imperial library had a clear-cut scope in collection—the original collation project in its commission only dealt with six categories of texts. Legal codes, for example, were entirely left out though they existed in large quantities and were especially important for governing at the time (Hulsewé, 1986). Thus, the catalog's classification should be more appropriately regarded as taking a holistic approach rather than being a comprehensive representation of the knowledge universe as perceived at the time.

A Classification with Correlative Thinking

Chinese philosophy is well-known for its correlative thinking which is closely associated with the holistic worldview. Renowned sinologist Joseph Needham describes that in this tradition “phenomena were thought of as parts of a hierarchy of wholes forming a cosmic pattern in which everything acted on everything else, not by mechanical impulsions but by co-operation in accord with the spontaneous motivations of its own inner nature” (2004, p. 91). Psychologists characterize this as “associative thinking” and consider it opposite of “analytic thinking” (Sloman, 2002). While the western tradition privileges the latter, it is generally recognized that the former dominates Chinese thought (Ames, 2002; Buchtel & Norenzayan, 2009). Applied to Chinese knowledge organization, this associative or correlative thinking gives overwhelming emphasis to the interrelationships among categories and the context.

In the classification of the Seven Epitomes, the focus of knowledge is the joint wisdom of the Six Classics and the main relationships of concern are the ones between the Classics and other textual categories. While the six main classes of texts are not his invention, Liu Xin carefully elucidates his personal interpretation of each category through a description that is expressive and analogous, never analytic. Here is a sample of such a description, somewhat abbreviated:

The masters form ten traditions, of which nine can be observed. They all rose from the fact that the Way of kings had grown obscure … Although their [i.e., the masters'] words differed, like fire and water they both destroyed and gave birth to one another … Although they are obstructed or weak, if you join their essential conclusions they are all branches or channels of the Six Canons. If their followers encounter an enlightened king or sage ruler who finds their common points, then they all have the ability to serve as his limbs. (The introductory summary for the Epitome of the Masters, translated by Lewis 1999, pp. 328–329)

This description contains no precise definition of the epitome or the individual masters and no comparison of the characteristics or theses of the various intellectual traditions (meaning the groups of the masters). On the other hand, it informs the reader of how the masters' writings relate to the Six Classics (“Six Canons” in the above translation by Lewis).

Citing linguistic theory of the late 20th century, Graham (1989) maintains that the tendency of Chinese thought, as demonstrated in the Chinese language, is to think in terms of whole/part rather than class/member relationships. That is, the parts of a whole are considered in terms of their relationships with the whole, not their similarities to/differences from one another. In the above quote, all the masters are likened to be the sage ruler's limbs. It is not to say that the sage ruler has dozens of limbs anatomically. Instead, a limb is only an analogy of a part and the masters' writings are all simply parts of the supreme wisdom regardless of whether they share a common set of characteristics.

In their tendency to think correlatively, the Chinese are not inclined to perceive objects, concepts, or events individually, detached from each other and the context (Buchtel & Norenzayan, 2009). Knowledge organization in the Seven Epitomes clearly demonstrates a lack of the kind of deductive analysis so commonplace in Dewey Decimal Classification and other widely used modern-day bibliographic classifications. A couple of advanced mathematical texts, for example, are in the subclass of Chronology, Epitome of Divination and Numbers, because they are a tool used in calendar making.44 Nisbett (2003) calls this kind of relationship in Chinese thinking “a thematic relationship”, contrasting it to a taxonomic relationship typical of western classification. Instances of thematic relationships are abundant in the Seven Epitomes. Unfortunately, such examples are precisely the ones that draw criticisms disparaging the scheme as being illogical, unscientific and wrong (Yao, 1937/2005; Tan, 2003). In effect, it is the critics who fail to acknowledge or appreciate the correlative approach to classification.


The current paper identified and discussed an alternative to the western analytic approach to classification—the traditional Chinese approach inaugurated and exemplified by the scheme in the Seven Epitomes. As seen above, the target classification purposefully embraces the ru classicist ideology as its guiding principle, presents a holistic conception of the knowledge universe, and applies correlative thinking to organizing all textual categories by placing the ru Classics at the center of all major relationships. Such a scheme at its face value may seem alien or even preposterous. However its fundamental ideas and methods provide instructional lessons for us to re-evaluate what we have been accustomed to. The fact that this classificatory model dominated and persisted in China for 2,000 years invalidates the universality of the western analytic model.

Further research may take several directions. For one possibility, the investigation of the traditional Chinese model of knowledge organization needs to go beyond merely a simplistic and superficial understanding of a few aspects as those reported in the current paper. It will also be profitable to examine epistemological approaches taken by other schemes in later times. Another potentially worthy pursuit is to test how the holistic and correlative thinking can be incorporated into knowledge organization based on modern advanced technology and how it can work in constantly evolving information environments. In addition, research must look at categories and relationships in their cultural contexts since they are socially constructed and have real social functions and consequences. These suggestions reflect my personal ideal of information services built with cultural sensitivity and accommodation for individual information-seeking and learning styles.


Dr. Wen-Chin Lan has been an invaluable contributor to this ongoing research project. I also wish to thank Dr. Jens-Erik Mai for his constructive suggestions.


  1. 1

    In the paper, all authors of Chinese works are presented with the family name first.

  2. 2

    For this reason, the Seven Epitomes and the Han Treatise are often regarded as the same work. In this section, I use the former to refer to the work. Readers should bear in mind that the Han Treatise might be the title referred to in certain cited publications.

  3. 3

    The exact nature of the Collective Epitome is unknown. Most scholars believe that it is a collection of the Preface and introductory summaries of individual epitomes and their subdivisions. In the Han Treatise, these segments are scattered.

  4. 4

    Texts on elementary mathematics, however, seem to be intentionally excluded from the Seven Epitomes. For a more detailed discussion, see Lee (in press).